drepr.models.parsers.v1.align_parser

Classes

AlignParser()

Alignments are defined as a list

class AlignParser

Bases: object

Alignments are defined as a list

``` - type: <alignment_type>

# other properties

```

If <alignment_type> is dimension, we have two schemas:

1. align by one dimension

``` - type: dimension

value: <source_attr_id>:<step_index> <-> <target_attr_id>:<step_index>

```

2. align by multiple dimensions

``` - type: dimension

source: <attr_id> target: <attr_id> aligned_dims:

• { source: <step_index>, target: <step_index> }

• { source: <step_index>, target: <step_index> }

…

```

If <alignment_type> is value, then the schema is

``` - type: value

source: <attr_id> target: <attr_id>

```

If <alignment_type> is auto, then the schema is:

``` - type: auto

attributes: <list of attributes to align>

```

The attributes are optional, if not provided, the system will try to align all attributes automatically based on the dimension

ALIGNMENT_TYPES_VALUES = {'auto', 'dimension', 'identical', 'value'}

REG_DALIGN = re.compile('^([^:]+)(?::(\\d+))? *<-> *([^:]+)(?::(\\d+))?$')

DALIGN_KEYS = {'aligned_dims', 'source', 'target', 'type'}

DALIGN_ALIGNED_DIMS_KEYS = {'source', 'target'}

VALIGN_KEYS = {'source', 'target', 'type'}

classmethod parse(alignments: List[dict]) → List[Union[RangeAlignment, ValueAlignment, IdenticalAlign, AutoAlignment]]

Parameters:

alignments (List[dict]) –

Return type:

List[Union[RangeAlignment, ValueAlignment, IdenticalAlign, AutoAlignment]]

classmethod parse_range_align(conf: dict, parse_trace: str) → RangeAlignment

Parameters:

• conf (dict) –

• parse_trace (str) –

Return type:

RangeAlignment

classmethod parse_value_align(conf: dict, parse_trace: str) → ValueAlignment

Parameters:

• conf (dict) –

• parse_trace (str) –

Return type:

ValueAlignment

classmethod parse_auto_align(conf: dict, parse_trace: str) → AutoAlignment

Parameters:

• conf (dict) –

• parse_trace (str) –

Return type:

AutoAlignment